This is really neat. I've recently been watching people use custom pi harnesses and thinking of jumping in, but I was dreaming of a way to do essentially what pi-msg does so that I could drive it remotely in a more full-featured way. Thanks for sharing.

Thanks! I had the same urge, I didn't want to be responsible for an entire GUI and communication layer but I did want to have a remotely controllable extensible harness, this was the perfect middle ground.

The benefit beyond that was having agents running on my server directly so they can do a lot of sysadmin and monitoring tasks, as well as suggesting config changes through PRs to my nix config.

Letting the agents bang their heads against the wall hunting down a config fix for a random malfunctioning piece of my homelab instead of me is by far my favorite part of the AI present.

The way I understand it, the connection is negotiated via BT, but then wifi is used for the fat data pipe to run the display.

The comments on this thread point to not very many people being aware of this.

I couldn't think of a use case for something this. For anyone else like me, the readme outlines one near the bottom which seems genuinely nice:

Set this up, allow no-auth sending from any address on your LAN, and then spinning up new local services that need outgoing mail is easy, because you can just give them the IP of the relay and omit credentials.

Exactly - the homelab case is the most common one. Instead of putting Gmail app passwords into 10 different devices, you configure it once in SimpleRelay and point everything at one endpoint.

At acme corp we used to spoof bonus emails from management through the smtp relay. Good times. Not so good security.

I am constantly surprised how seldom aider is mentioned in threads like this. I understand that it's not directly integrated into the editor, but the "editor + parallel CLI tool chain" paradigm feels so natural to me because we drop to terminal for so many other parts of building software. If you haven't tried it (particularly the architect/editor modality), it's worth a couple of hours of experimenting.

Aider doesn’t provide any interface that’s integrated into the editor tool, as you point out. That might be true for other similar side-by-side tools that I am not aware of.

But, if you tell aider to watch your files, you can drop a specially formatted comment into your file, and aider will see that and use it as a prompt.

So the integration is sort of “implicit”. Which sounds kinda like the cheap way to go, in comparison to the current brand name tools that have their own chat boxes, dropdowns with mode selectors (ask, edit, agent), and so on.

But look further into the future and an ambient interface is probably where we end up. Something where the Ai agent is just watching what you do, maybe even watching your eyes and seeing what you’re attending to, and then harmonizing its attention to what you are attending to.

But I dunno, i’m just guessing

Tight editor integration means better diffs (right in your editor), easier context manager, and other convenience features that CLI-only tools can't have.

This doesn't mean that aider, claude code, etc. aren't very good tools, but it does make sense to distinguish between built-in tools vs external ones. A similar non-AI example is debugging or linting: IDE integration makes it much easier than using a separate tool.

I like to imagine as if we threw something into the air so hard that it went up for 26 years, then came back down for 26 years.

That's exactly what happened, except it went around the world a million times too.

The best part about Kagi is that if the default results don't seem helpful, one click restricts results to only discussions and forums, which is usually exactly what I want to do next.

Google supports that too. After searching, click “Forums” in the top bar.

As my income drops in slow periods, I pause subscriptions to many things. The very last two I would be willing to give up are (n-1): YouTube Premium and (n): Kagi.

Switched to nebula when they removed YouTube lite and the price was double. Got no ads on my devices anyway.

For me, YouTube Premium (for family) is so worth it because:

- It removes ads for my kids, which are going to watch YouTube no matter what I do

- It includes YouTube Music, so I don't need a separate Spotify subscription

- It sends money to creators based on what I proportionally spend time watching, rather than based on how valuable of an audience I'm deemed to be when an ad slot is auctioned

- It is a signal that I'm willing to pay for content, and I want platforms and creators to continue offering that option

Rather send money directly to creators rather than filtering it through Google.

People just don't do that. If they did, the state of things wouldn't be the way it is. It's also kinda unfair to creators whose content people consume but just don't pay for in any way, because of a variety of possible reasons (they just don't even think about it, or they don't deem them "worthy" of being paid, or it's about having to go out of their way to pay someone, and then people only paying a limited amount of "chosen" people, etc.)

That also is a factor in why I pay for more than I need. I want them to succeed.

Personally, I would still love a site-by-site "reject non-essential cookies" prompt from an extension that's in the same place, with the same UI, on every site. Still a click, but lots better than having to figure out how to accomplish it on each and every site.

Exactly. The biggest pain is to read and figure out what the next button actually does. Is the big Button an except all? Use selected? Or what ever wording they use. I might not want to block cookies for certain pages. So an extension that finally creates this single UX flow would be very helpful indeed.

Exactly. So you could have 2 shortcuts: one for reject all non essential, one for accept all.

Much better UX than figuring out per site which button to click.

Why would you ever accept all? The options should be reject all non essential and reject all (may break something)